Any other business
The Memrise Prize: A project to create an olympics for learning science
7 min read
09 February 2017
While some projects get confined to the waste bin, others bubble in the background. For Ed Cooke, the Memrise Prize has been a slow, but eventual, burner.
The unveiling of our Memrise Prize is evidence that it’s important to keep hold of ideas and initiatives you feel passionately about – even as the sands of time shift around you.
One of the interesting challenges of any complex enterprise is that the projects that comprise it take different amounts of time.
Most are fairly short and immediately goal-oriented. Others take months and need to be managed and calibrated with special foresight. And every so often, you have a project which, like a tree in a garden, comes to maturity years later than the flowers and shrubs that surround it – sometimes long after the person who planted has forgotten having done so.
Such projects are rare, for they’re normally orthogonal to business goals. But when they do occur, they often shimmer with a magic blossom.
One such project, planted in a different era of Memrise’s evolution nearly three years ago, is happily reaching maturity this month. When we conceived the idea we were six people running a user-generated website with no revenue at all. We’re now 45 people with profitable mobile apps and tens of millions of users.
Almost everything has changed in that time, so it’s remarkable the project has survived. It’s also fortunate too, as it’s a magic and nutty idea and it might just be the most interesting thing we’ve ever been a part of.
The project is called the Memrise Prize. It’s a bit like an Olympics for learning methods. It emerged from the desire, shared by us and scientists around the world, to stimulate a more profound and useful science of learning; to answer the question “what’s the best way to learn?”. Amazingly, there’s been very little research directly approaching this question.
In order to ask the question truly scientifically, which meant with objectivity, conceptual clarity and radical open-ness, we leant upon David Shanks and Rosalind Potts at UCL, some of the world’s most distinguished memory scientists, to help design an open experimental framework within which anyone could create a learning methodology and compare it with anyone else’s method for effectiveness. A system in which scientists from round the world could invent learning methods and race them.
Who’s the fastest man in the world? Before the modern Olympics were founded 120 years ago, nobody would have had any idea. With the Olympics – an open competition in which every nation can put forward their finest athletes – we have a system that can reliably identify the fastest: Usain Bolt, as it turned out, the last few times. With the Memrise Prize, we sought to invent a system of similar lucidity, but for learning science. A competition for identifying the speediest learning methods to breed innovation in learning science.
With this in mind we invited scientists, students and, in fact, anyone who wished to compete from around the world to create the best method they could think of that would help a population of internet users learn 80 foreign words in an hour, with a test one week later.
Teams from many of the world’s top cognitive science departments competed in the Memrise Prize, among them Oxford, MIT, Stanford, Washington State. From Europe, wonderful international conglomerates of scientists came together to pool their expertise to interactive learning methods – drawing from diverse ideas from the history of learning techniques and science.
With these wonderfully inventive (and very diverse) learning methods in hand, we created them on our experimental system and then ran (with the extraordinary generosity and help of the Memrise community) one of the largest psychology experiments ever conducted, in which tens of thousands of people tested these methods and tried to remember as much as they could. The method which helped people learn the most one week after studying would win.
So anyhow, after three years of work, the results are all in from our gigantic scientific experiment and our international Memrise Prize panel of judges are poring over the data to anoint the winner.
We can’t announce that winner now, but we can’t wait to do so in the coming months, as well as reuniting them with their $10,000 prize. What we can say about several of the top methodologies is that they show a more than doubling of effectiveness against flashcard methods (which are already much more powerful than self-study) and so represent a huge leap in the possibilities of the practical application of scientific knowledge to actual learning. Learners around the world stand to benefit.
We’ll be enthusiastic in how we integrate the learnings into our own learning methods, and because the results are open (according to the spirit of science), it will give all the many learning technology startups in search of an effective learning method the chance to find one.
And for our old colleague Eliot York, who left Memrise a few years ago to run his poetry website hellopoetry.com from a mountain in New Mexico, it represents the completion of a project that joins the past to the present: like memory itself.
And just as memory constantly renews, we aim for the Memrise Prize to become an annual event, and so help stimulate a continuous honing and evolution of learning methods over time, something that is long overdue for our society.